You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@hbase.apache.org by Slava Gorelik <sl...@gmail.com> on 2008/11/04 09:26:37 UTC
Re: Regionserver fails to serve region
Hi Michael.After reformatting HDFS, Hbase started to work as a Swiss Clock.
Worked with 8 clients about 30 hours intensive load.
Just small question, after about 28 hours (when i came back to work) i found
that one of 7 datanodes in Hadoop is about 98% usage and all other about
30%, is it normal ?
Best Regards.
On Fri, Oct 31, 2008 at 10:16 PM, Slava Gorelik <sl...@gmail.com>wrote:
> Hi.No problem with silly question :-) Yes, sure i replaced, here the list
> of folder that begins with 73*:
>
> drwxr-xr-x - XXXXXXXXX supergroup 0 2008-10-29 11:13 /hbase/BizDB/732078971/BusinessObject
> drwxr-xr-x - XXXXXXXX supergroup 0 2008-10-29 11:13 /hbase/BizDB/732215319/BusinessObject
> drwxr-xr-x - XXXXXXXX supergroup 0 2008-10-29 11:13 /hbase/BizDB/733411255/BusinessObject
> drwxr-xr-x - XXXXXXXX supergroup 0 2008-10-29 11:14 /hbase/BizDB/733598097/BusinessObject
> drwxr-xr-x - XXXXXXXX supergroup 0 2008-10-29 10:50 /hbase/BizDB/734145833/BusinessObject
> drwxr-xr-x - XXXXXXXX supergroup 0 2008-10-29 11:09 /hbase/BizDB/735612900/BusinessObject
> drwxr-xr-x - XXXXXXXX supergroup 0 2008-10-29 11:15 /hbase/BizDB/738009120/BusinessObject
>
> There no 735893330 folder.
> Scanning.META. in shell is not easy at all. .META. is huge and simple scan
> without providing specific column will give me about 10 min only listing of
> .META. content, so i failed to find the 735893330, may be you can give me
> the name of the column, where this info is placed ?
>
> I think i'll reformat the HDFS and will start it from clean environment and
> the we'll see. I'll do it this Sunday and let you know.
>
> Best Regards and Big Thank You for your patience and assistance.
>
>
> On Fri, Oct 31, 2008 at 4:47 AM, Michael Stack <st...@duboce.net> wrote:
>
>> Slava Gorelik wrote:
>>
>>> Hi.I also noticed this exception.
>>> Strange that this exception is happened every time on the same
>>> regionserver.
>>> Tried to find directory hdfs://X:9000/hbase/BizDB/735893330 - not exist.
>>> Very strange, but history folder in hadoop is empty.
>>>
>>>
>> It is odd indeed that the system keeps trying to load a region that does
>> not exist.
>>
>> I don't think it necessarily the same regionserver that is responsible.
>> I'd think it an attribute of the region that we're trying to deploy on that
>> server.
>>
>> Silly question: you did replace 'X' with your machine name in the above?
>>
>> If you restart, it still tries to load this nonexistent region?
>>
>> If so, the .META. table is not consistent with whats on the filesystem.
>> They've gotten out of sync. Describing how to repair is involved.
>>
>> Reformatting HDFS will help ?
>>>
>>>
>>>
>> Do a "scan '.META.'" in the shell. Do you see your region listed (look at
>> the encoded names attribute to find 735893330.
>>
>> If your table is damaged -- i'd guess it because ulimit was bad up to this
>> -- the best thing might to start over.
>>
>> One more things in a last minute, i found that one node in cluster has
>>> totally different time, could this cause for such a problems ?
>>>
>>>
>> We thought we'd fixed all problems that could arise from time skew, but
>> you never know. In our requirements, clocks must be synced. Fix this too
>> if you can before reloading.
>>
>> P.S. About logs, is it possible to send to some email - each log file
>>> compressed is about 1mb, and only in 3 files i found exceptions.
>>>
>>>
>>>
>> There probably is such a functionality but I'm not familiar. Can you put
>> them under a webserver at your place so I can grab them? You can send me
>> the URL offlist if you like.
>>
>> Thanks for your patience Slava. We'll figure it.
>>
>> St.Ack
>>
>>
>> On Thu, Oct 30, 2008 at 10:25 PM, stack <st...@duboce.net> wrote:
>>>
>>>
>>>
>>>> Can you put them someplace that I can pull them?
>>>>
>>>> I took another look at your logs. I see that a region is missing files.
>>>> That means it will never open and just keep trying. Grep your logs for
>>>> FileNotFound. You'll see this:
>>>>
>>>>
>>>> hbase-clmanager-regionserver-ILREDHAT012.log:java.io.FileNotFoundException:
>>>> File does not exist:
>>>>
>>>> hdfs://X:9000/hbase/BizDB/735893330/BusinessObject/mapfiles/647541142630058906/data
>>>>
>>>> hbase-clmanager-regionserver-ILREDHAT012.log:java.io.FileNotFoundException:
>>>> File does not exist:
>>>>
>>>> hdfs://X:9000/hbase/BizDB/735893330/BusinessObject/mapfiles/2243545870343537637/data
>>>>
>>>> Try shutting down, and removing these files. Remove the following
>>>> directories:
>>>>
>>>>
>>>>
>>>> hdfs://X:9000/hbase/BizDB/735893330/BusinessObject/mapfiles/647541142630058906
>>>>
>>>> hdfs://X:9000/hbase/BizDB/735893330/BusinessObject/info/647541142630058906
>>>>
>>>>
>>>> hdfs://X:9000/hbase/BizDB/735893330/BusinessObject/mapfiles/2243545870343537637
>>>>
>>>> hdfs://X:9000/hbase/BizDB/735893330/BusinessObject/info/2243545870343537637
>>>>
>>>> Then retry restarting.
>>>>
>>>> You can try and figure how these files got lost by going back in your
>>>> history.
>>>>
>>>>
>>>> St.Ack
>>>>
>>>>
>>>>
>>>> Slava Gorelik wrote:
>>>>
>>>>
>>>>
>>>>> Michael,still have the problem, but the logs files are very big (50MB
>>>>> each)
>>>>> even compressed they are bigger than limit for this mailing list.
>>>>> Most of the problems are happened during compaction (i see in the log),
>>>>> may
>>>>> be i can send some parts from logs ?
>>>>>
>>>>> Best Regards.
>>>>>
>>>>> On Thu, Oct 30, 2008 at 8:49 PM, Slava Gorelik <
>>>>> slava.gorelik@gmail.com
>>>>>
>>>>>
>>>>>> wrote:
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>>> Sorry, my mistake, i did it for wrong user name.Thanks, updating now,
>>>>>> soon
>>>>>> will try again.
>>>>>>
>>>>>>
>>>>>> On Thu, Oct 30, 2008 at 8:39 PM, Slava Gorelik <
>>>>>> slava.gorelik@gmail.com
>>>>>>
>>>>>>
>>>>>>> wrote:
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>> Hi.Very strange, i see in limits.conf that it's upped.
>>>>>>> I attached the limits.conf, please have a look, may be i did it
>>>>>>> wrong.
>>>>>>>
>>>>>>> Best Regards.
>>>>>>>
>>>>>>>
>>>>>>> On Thu, Oct 30, 2008 at 7:52 PM, stack <st...@duboce.net> wrote:
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>> Thanks for the logs Slava. I notice that you have not upped the
>>>>>>>> ulimit
>>>>>>>> on your cluster. See the head of your logs where we print out the
>>>>>>>> ulimit.
>>>>>>>> Its 1024. This could be one cause of your grief especially when
>>>>>>>> you
>>>>>>>> seemingly have many regions (>1000). Please try upping it.
>>>>>>>> St.Ack
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Slava Gorelik wrote:
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>> Hi.
>>>>>>>>> I enabled DEBUG log level and now I'm sending all logs (archived)
>>>>>>>>> including fsck run result.
>>>>>>>>> Today my program starting to fail couple of minutes from the begin,
>>>>>>>>> it's
>>>>>>>>> very easy to reproduce the problem, cluster became very unstable.
>>>>>>>>>
>>>>>>>>> Best Regards.
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> On Tue, Oct 28, 2008 at 11:05 PM, stack <stack@duboce.net <mailto:
>>>>>>>>> stack@duboce.net>> wrote:
>>>>>>>>>
>>>>>>>>> See http://wiki.apache.org/hadoop/Hbase/FAQ#5
>>>>>>>>>
>>>>>>>>> St.Ack
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Slava Gorelik wrote:
>>>>>>>>>
>>>>>>>>> Hi.First of all i want to say thank you for you assistance !!!
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> DEBUG on hadoop or hbase ? And how can i enable ?
>>>>>>>>> fsck said that HDFS is healthy.
>>>>>>>>>
>>>>>>>>> Best Regards and Thank You
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> On Tue, Oct 28, 2008 at 8:45 PM, stack <stack@duboce.net
>>>>>>>>> <ma...@duboce.net>> wrote:
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Slava Gorelik wrote:
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Hi.HDFS capacity is about 800gb (8 datanodes) and the
>>>>>>>>> current usage is
>>>>>>>>> about
>>>>>>>>> 30GB. This is after total re-format of the HDFS that
>>>>>>>>> was made a hour
>>>>>>>>> before.
>>>>>>>>>
>>>>>>>>> BTW, the logs i sent are from the first exception that
>>>>>>>>> i found in them.
>>>>>>>>> Best Regards.
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Please enable DEBUG and retry. Send me all logs. What
>>>>>>>>> does the fsck on
>>>>>>>>> HDFS say? There is something seriously wrong with your
>>>>>>>>> cluster that you are
>>>>>>>>> having so much trouble getting it running. Lets try and
>>>>>>>>> figure it.
>>>>>>>>>
>>>>>>>>> St.Ack
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> On Tue, Oct 28, 2008 at 7:12 PM, stack
>>>>>>>>> <stack@duboce.net <ma...@duboce.net>> wrote:
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> I took a quick look Slava (Thanks for sending the
>>>>>>>>> files). Here's a few
>>>>>>>>> notes:
>>>>>>>>>
>>>>>>>>> + The logs are from after the damage is done; the
>>>>>>>>> transition from good to
>>>>>>>>> bad is missing. If I could see that, that would
>>>>>>>>> help
>>>>>>>>> + But what seems to be plain is that that your
>>>>>>>>> HDFS is very sick. See
>>>>>>>>> this
>>>>>>>>> from head of one of the regionserver logs:
>>>>>>>>>
>>>>>>>>> 2008-10-27 23:41:12,682 WARN
>>>>>>>>> org.apache.hadoop.dfs.DFSClient:
>>>>>>>>> DataStreamer
>>>>>>>>> Exception: java.io.IOException: Unable to create
>>>>>>>>> new block.
>>>>>>>>> at
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> org.apache.hadoop.dfs.DFSClient$DFSOutputStream.nextBlockOutputStream(DFSClient.java:2349)
>>>>>>>>> at
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> org.apache.hadoop.dfs.DFSClient$DFSOutputStream.access$1800(DFSClient.java:1735)
>>>>>>>>> at
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> org.apache.hadoop.dfs.DFSClient$DFSOutputStream$DataStreamer.run(DFSClient.java:1912)
>>>>>>>>>
>>>>>>>>> 2008-10-27 23:41:12,682 WARN
>>>>>>>>> org.apache.hadoop.dfs.DFSClient: Error
>>>>>>>>> Recovery for block blk_-5188192041705782716_60000
>>>>>>>>> bad datanode[0]
>>>>>>>>> 2008-10-27 23:41:12,685 ERROR
>>>>>>>>>
>>>>>>>>> org.apache.hadoop.hbase.regionserver.CompactSplitThread:
>>>>>>>>> Compaction/Split
>>>>>>>>> failed for region
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> BizDB,1.1.PerfBO1.f2188a42-5eb7-4a6a-82ef-2da0d0ea4ce0,1225136351518
>>>>>>>>> java.io.IOException: Could not get block
>>>>>>>>> locations. Aborting...
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> If HDFS is ailing, hbase is too. In fact, the
>>>>>>>>> regionservers will shut
>>>>>>>>> themselves to protect themselves against damaging
>>>>>>>>> or losing data:
>>>>>>>>>
>>>>>>>>> 2008-10-27 23:41:12,688 FATAL
>>>>>>>>> org.apache.hadoop.hbase.regionserver.Flusher:
>>>>>>>>> Replay of hlog required. Forcing server restart
>>>>>>>>>
>>>>>>>>> So, whats up with your HDFS? Not enough space
>>>>>>>>> alloted? What happens if
>>>>>>>>> you run "./bin/hadoop fsck /"? Does that give you
>>>>>>>>> a clue as to what
>>>>>>>>> happened? Dig in the datanode and namenode logs.
>>>>>>>>> Look for where the
>>>>>>>>> exceptions start. It might give you a clue.
>>>>>>>>>
>>>>>>>>> + The suse regionserver log had garbage in it.
>>>>>>>>>
>>>>>>>>> St.Ack
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Slava Gorelik wrote:
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Hi.
>>>>>>>>> My happiness was very short :-( After i
>>>>>>>>> successfully added 1M rows (50k
>>>>>>>>> each row) i tried to add 10M rows.
>>>>>>>>> And after 3-4 working hours it started to
>>>>>>>>> dying. First one region server
>>>>>>>>> is died, after another one and eventually all
>>>>>>>>> cluster is dead.
>>>>>>>>>
>>>>>>>>> I attached log files (relevant part, archived)
>>>>>>>>> from region servers and
>>>>>>>>> from the master.
>>>>>>>>>
>>>>>>>>> Best Regards.
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> On Mon, Oct 27, 2008 at 11:19 AM, Slava
>>>>>>>>> Gorelik
>>>>>>>>> <
>>>>>>>>> slava.gorelik@gmail.com
>>>>>>>>> <ma...@gmail.com><mailto:
>>>>>>>>> slava.gorelik@gmail.com
>>>>>>>>> <ma...@gmail.com>>> wrote:
>>>>>>>>>
>>>>>>>>> Hi.
>>>>>>>>> So far so good, after changing the file
>>>>>>>>> descriptors
>>>>>>>>> and dfs.datanode.socket.write.timeout,
>>>>>>>>> dfs.datanode.max.xcievers
>>>>>>>>> my cluster works stable.
>>>>>>>>> Thank You and Best Regards.
>>>>>>>>>
>>>>>>>>> P.S. Regarding deleting multiple columns
>>>>>>>>> missing functionality i
>>>>>>>>> filled jira :
>>>>>>>>>
>>>>>>>>> https://issues.apache.org/jira/browse/HBASE-961
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> On Sun, Oct 26, 2008 at 12:58 AM, Michael
>>>>>>>>> Stack <stack@duboce.net <mailto:
>>>>>>>>> stack@duboce.net
>>>>>>>>> <mailto:stack@duboce.net
>>>>>>>>>
>>>>>>>>> <ma...@duboce.net>>> wrote:
>>>>>>>>>
>>>>>>>>> Slava Gorelik wrote:
>>>>>>>>>
>>>>>>>>> Hi.Haven't tried yet them, i'll try
>>>>>>>>> tomorrow morning. In
>>>>>>>>> general cluster is
>>>>>>>>> working well, the problems begins if
>>>>>>>>> i'm trying to add 10M
>>>>>>>>> rows, after 1.2M
>>>>>>>>> if happened.
>>>>>>>>>
>>>>>>>>> Anything else running beside the
>>>>>>>>> regionserver or datanodes
>>>>>>>>> that would suck resources? When
>>>>>>>>> datanodes begin to slow, we
>>>>>>>>> begin to see the issue Jean-Adrien's
>>>>>>>>> configurations address.
>>>>>>>>> Are you uploading using MapReduce? Are
>>>>>>>>> TTs running on same
>>>>>>>>> nodes as the datanode and regionserver?
>>>>>>>>> How are you doing the
>>>>>>>>> upload? Describe what your uploader
>>>>>>>>> looks like (Sorry if
>>>>>>>>> you've already done this).
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> I already changed the limit of files
>>>>>>>>> descriptors,
>>>>>>>>>
>>>>>>>>> Good.
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> I'll try
>>>>>>>>> to change the properties:
>>>>>>>>> <property>
>>>>>>>>> <name>dfs.datanode.socket.write.timeout</name>
>>>>>>>>> <value>0</value>
>>>>>>>>> </property>
>>>>>>>>>
>>>>>>>>> <property>
>>>>>>>>>
>>>>>>>>> <name>dfs.datanode.max.xcievers</name>
>>>>>>>>> <value>1023</value>
>>>>>>>>> </property>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Yeah, try it.
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> And let you know, is any other
>>>>>>>>> prescriptions ? Did i miss
>>>>>>>>> something ?
>>>>>>>>>
>>>>>>>>> BTW, off topic, but i sent e-mail
>>>>>>>>> recently to the list and
>>>>>>>>> i can't see it:
>>>>>>>>> Is it possible to delete multiple
>>>>>>>>> columns in any way by
>>>>>>>>> regex : for example
>>>>>>>>> colum_name_* ?
>>>>>>>>>
>>>>>>>>> Not that I know of. If its not in the
>>>>>>>>> API, it should be.
>>>>>>>>> Mind filing a JIRA?
>>>>>>>>>
>>>>>>>>> Thanks Slava.
>>>>>>>>> St.Ack
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>>>
>>
>>
>
Re: Regionserver fails to serve region
Posted by Slava Gorelik <sl...@gmail.com>.
Hi.
I'll enable DEBUG and run cluster this weekend , i need it stable till
weekend.
But, as a Murphy's Law it will not happen with DEBUG on :-).
BTW, there no reason to run balancer on the beginning, from last time
balancer was required after 30h work.
After 30h of work - log files are large than it should be :-)
I'll run it and let you know after weekend.
Thank You for your assistance an patience.
On Tue, Nov 11, 2008 at 6:47 PM, Michael Stack <st...@duboce.net> wrote:
> Slava Gorelik wrote:
>
>> Hi.DEBUG wasn't enabled , because it decrease the performance and increase
>> log size.
>>
>>
> Sure. But maybe leave it on while we're trying to figure issues.
>
> Regarding the ulimit - yes it's upped for 32K.
>>
>>
> Good.
>
> You remember correct - during massive load i run the balancer and from
>> this
>> time everything is started to behave strange.
>> Currently , i can't tell you the the regions that are in the table - i
>> re-formatted hdfs ( this was the only way i can get my cluster back to
>> work).
>>
>>
> Sure (If DEBUG was on, it records in log how many -- just FYI).
>
> I have 7 datatnodes , 6 of them are running region server and one is
>> Hmaster.
>>
>>
> Are things running for you now? Have you tried another upload without the
> balancer?
> St.Ack
>
Re: Regionserver fails to serve region
Posted by Michael Stack <st...@duboce.net>.
Slava Gorelik wrote:
> Hi.DEBUG wasn't enabled , because it decrease the performance and increase
> log size.
>
Sure. But maybe leave it on while we're trying to figure issues.
> Regarding the ulimit - yes it's upped for 32K.
>
Good.
> You remember correct - during massive load i run the balancer and from this
> time everything is started to behave strange.
>
> Currently , i can't tell you the the regions that are in the table - i
> re-formatted hdfs ( this was the only way i can get my cluster back to
> work).
>
Sure (If DEBUG was on, it records in log how many -- just FYI).
> I have 7 datatnodes , 6 of them are running region server and one is
> Hmaster.
>
Are things running for you now? Have you tried another upload without
the balancer?
St.Ack
Re: Regionserver fails to serve region
Posted by Slava Gorelik <sl...@gmail.com>.
Hi.DEBUG wasn't enabled , because it decrease the performance and increase
log size.
Regarding the ulimit - yes it's upped for 32K.
You remember correct - during massive load i run the balancer and from this
time everything is started to behave strange.
Currently , i can't tell you the the regions that are in the table - i
re-formatted hdfs ( this was the only way i can get my cluster back to
work).
I have 7 datatnodes , 6 of them are running region server and one is
Hmaster.
Best Regards.
On Tue, Nov 11, 2008 at 1:08 AM, stack <st...@duboce.net> wrote:
> I took a look.
>
> First, enable DEBUG. See the hbase FAQ for how.
>
> Looking, I see that all was running fine till:
>
> 2008-11-03 14:10:08,261 INFO org.apache.hadoop.ipc.Client: Retrying connect
> to server: /10.X.X.Y:60020. Already tried 0 time(s).
>
> ...in the middle of an attempt at scanning the .META. region.
>
> Looking through regionserver logs, they are all fine till about that above
> time when I start to see variations on:
>
> 2008-11-03 14:08:46,440 INFO org.apache.hadoop.dfs.DFSClient: Could not
> obtain block blk_1223341017118968735_305051 from any node:
> java.io.IOException: No live nodes contain current block
>
> ....and
>
> 2008-11-03 14:08:43,660 INFO org.apache.hadoop.dfs.DFSClient: Exception in
> createBlockOutputStream java.io.IOException: Bad connect ack with
> firstBadLink 10.X.X.Y:50010
> 2008-11-03 14:08:43,660 INFO org.apache.hadoop.dfs.DFSClient: Abandoning
> block blk_6726606309673852040_314096
>
> Your hdfs went bad for some reason around above time. I don't see any
> obvious explanation for why it went bad. You were running balancer at the
> time IIRC?
>
> Could you netstat your running datanodes and see how many concurrent
> connections you had running? Was 1024 enough? You had configured a max of
> 1024? I don't see the ulimit print out in these logs so presume its > 1024.
>
> How many regions do you have in your table when it starts to go wonky? You
> have 6 datanodes running beside your 6 regionservers?
>
> St.Ack
>
>
> Slava Gorelik wrote:
>
>> Hi Michael.
>> I'm sending logs, in 2 parts (2 messages)
>> Part 1
>>
>>
>> On Tue, Nov 4, 2008 at 11:44 PM, Slava Gorelik <slava.gorelik@gmail.com<mailto:
>> slava.gorelik@gmail.com>> wrote:
>>
>> Thank You. Now it's clear.
>>
>>
>> On Tue, Nov 4, 2008 at 11:31 PM, stack <stack@duboce.net
>> <ma...@duboce.net>> wrote:
>>
>> Slava Gorelik wrote:
>>
>> One more regarding the blockCache, how changes in store
>> files (as i
>> understand those are MapFiles) are reflected on client
>> side cache. If we are
>> talking about more than one client that doing a changes ?
>> If each client has
>> different part of the MapFile ? or something else ?
>>
>>
>> The block cache cache is over in the server. Its a cache for
>> store files which never change once written. Did I say
>> client-side cache? I should have been more clear. The client
>> in this case is the regionserver itself. The cache is so the
>> regionserver saves on its trips over the network visiting
>> datanodes.
>> St.Ack
>>
>>
>>
>> Best Regards.
>>
>> On Tue, Nov 4, 2008 at 11:10 PM, Slava Gorelik
>> <slava.gorelik@gmail.com
>> <ma...@gmail.com>>wrote:
>>
>>
>> I can try to reproduce it again, but before this i
>> would like to send you a
>> logs.
>> Best Regards.
>>
>>
>> On Tue, Nov 4, 2008 at 10:05 PM, stack
>> <stack@duboce.net <ma...@duboce.net>> wrote:
>>
>>
>> Then we should try and figure if there is an issue
>> in the balancer, or
>> maybe there is something missing if we are not
>> doing a big upload in a
>> manner that balances the upload across HDFS?
>> St.Ack
>>
>> Slava Gorelik wrote:
>>
>>
>> Sure, i'll arrange logs tomorrow.About
>> balancer, to wait when the massive
>> work is finished is good in testing
>> environment but in production it's
>> not
>> relevant :-)
>>
>> Best Regards.
>>
>> On Tue, Nov 4, 2008 at 9:48 PM, stack
>> <stack@duboce.net <ma...@duboce.net>>
>>
>> wrote:
>>
>>
>>
>>
>> Slava Gorelik wrote:
>>
>>
>>
>>
>> Hi.Regarding the failure of new block
>> creation - i failed to run hbase
>> till
>> i reformatted HDFS again.
>>
>>
>>
>>
>>
>> I'd be interested in the logs.
>>
>> I just wandering if hadoop re balancing
>> is necessary? Will it balance
>>
>>
>>
>> itself
>> ? As i understand hadoop balancer is
>> moving data between data nodes,
>> but
>> in
>> my case this is during massive (8
>> clients just adding a records - about
>> 400
>> requests for all region servers - 6).
>> So, is it good idea to run
>> balancer during heavy load ?
>>
>>
>>
>>
>>
>> I don't have sufficient experience running
>> the balancer. Perhaps wait
>> till
>> upload is done, then run it?
>>
>> St.Ack
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>
Re: Regionserver fails to serve region
Posted by stack <st...@duboce.net>.
I took a look.
First, enable DEBUG. See the hbase FAQ for how.
Looking, I see that all was running fine till:
2008-11-03 14:10:08,261 INFO org.apache.hadoop.ipc.Client: Retrying
connect to server: /10.X.X.Y:60020. Already tried 0 time(s).
...in the middle of an attempt at scanning the .META. region.
Looking through regionserver logs, they are all fine till about that
above time when I start to see variations on:
2008-11-03 14:08:46,440 INFO org.apache.hadoop.dfs.DFSClient: Could not
obtain block blk_1223341017118968735_305051 from any node:
java.io.IOException: No live nodes contain current block
....and
2008-11-03 14:08:43,660 INFO org.apache.hadoop.dfs.DFSClient: Exception
in createBlockOutputStream java.io.IOException: Bad connect ack with
firstBadLink 10.X.X.Y:50010
2008-11-03 14:08:43,660 INFO org.apache.hadoop.dfs.DFSClient: Abandoning
block blk_6726606309673852040_314096
Your hdfs went bad for some reason around above time. I don't see any
obvious explanation for why it went bad. You were running balancer at
the time IIRC?
Could you netstat your running datanodes and see how many concurrent
connections you had running? Was 1024 enough? You had configured a max
of 1024? I don't see the ulimit print out in these logs so presume its
> 1024.
How many regions do you have in your table when it starts to go wonky?
You have 6 datanodes running beside your 6 regionservers?
St.Ack
Slava Gorelik wrote:
> Hi Michael.
> I'm sending logs, in 2 parts (2 messages)
> Part 1
>
>
> On Tue, Nov 4, 2008 at 11:44 PM, Slava Gorelik
> <slava.gorelik@gmail.com <ma...@gmail.com>> wrote:
>
> Thank You. Now it's clear.
>
>
> On Tue, Nov 4, 2008 at 11:31 PM, stack <stack@duboce.net
> <ma...@duboce.net>> wrote:
>
> Slava Gorelik wrote:
>
> One more regarding the blockCache, how changes in store
> files (as i
> understand those are MapFiles) are reflected on client
> side cache. If we are
> talking about more than one client that doing a changes ?
> If each client has
> different part of the MapFile ? or something else ?
>
>
>
> The block cache cache is over in the server. Its a cache for
> store files which never change once written. Did I say
> client-side cache? I should have been more clear. The client
> in this case is the regionserver itself. The cache is so the
> regionserver saves on its trips over the network visiting
> datanodes.
> St.Ack
>
>
>
> Best Regards.
>
> On Tue, Nov 4, 2008 at 11:10 PM, Slava Gorelik
> <slava.gorelik@gmail.com
> <ma...@gmail.com>>wrote:
>
>
>
> I can try to reproduce it again, but before this i
> would like to send you a
> logs.
> Best Regards.
>
>
> On Tue, Nov 4, 2008 at 10:05 PM, stack
> <stack@duboce.net <ma...@duboce.net>> wrote:
>
>
>
> Then we should try and figure if there is an issue
> in the balancer, or
> maybe there is something missing if we are not
> doing a big upload in a
> manner that balances the upload across HDFS?
> St.Ack
>
> Slava Gorelik wrote:
>
>
>
> Sure, i'll arrange logs tomorrow.About
> balancer, to wait when the massive
> work is finished is good in testing
> environment but in production it's
> not
> relevant :-)
>
> Best Regards.
>
> On Tue, Nov 4, 2008 at 9:48 PM, stack
> <stack@duboce.net <ma...@duboce.net>>
> wrote:
>
>
>
>
>
> Slava Gorelik wrote:
>
>
>
>
>
> Hi.Regarding the failure of new block
> creation - i failed to run hbase
> till
> i reformatted HDFS again.
>
>
>
>
>
>
> I'd be interested in the logs.
>
> I just wandering if hadoop re balancing
> is necessary? Will it balance
>
>
>
>
> itself
> ? As i understand hadoop balancer is
> moving data between data nodes,
> but
> in
> my case this is during massive (8
> clients just adding a records - about
> 400
> requests for all region servers - 6).
> So, is it good idea to run
> balancer during heavy load ?
>
>
>
>
>
>
> I don't have sufficient experience running
> the balancer. Perhaps wait
> till
> upload is done, then run it?
>
> St.Ack
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
Re: Regionserver fails to serve region
Posted by Slava Gorelik <sl...@gmail.com>.
Hi Michael.I'm sending logs, in 2 parts (2 messages)
Part 1
On Tue, Nov 4, 2008 at 11:44 PM, Slava Gorelik <sl...@gmail.com>wrote:
> Thank You. Now it's clear.
>
>
> On Tue, Nov 4, 2008 at 11:31 PM, stack <st...@duboce.net> wrote:
>
>> Slava Gorelik wrote:
>>
>>> One more regarding the blockCache, how changes in store files (as i
>>> understand those are MapFiles) are reflected on client side cache. If we
>>> are
>>> talking about more than one client that doing a changes ? If each client
>>> has
>>> different part of the MapFile ? or something else ?
>>>
>>>
>>
>> The block cache cache is over in the server. Its a cache for store files
>> which never change once written. Did I say client-side cache? I should
>> have been more clear. The client in this case is the regionserver itself.
>> The cache is so the regionserver saves on its trips over the network
>> visiting datanodes.
>> St.Ack
>>
>>
>>
>> Best Regards.
>>>
>>> On Tue, Nov 4, 2008 at 11:10 PM, Slava Gorelik <slava.gorelik@gmail.com
>>> >wrote:
>>>
>>>
>>>
>>>> I can try to reproduce it again, but before this i would like to send
>>>> you a
>>>> logs.
>>>> Best Regards.
>>>>
>>>>
>>>> On Tue, Nov 4, 2008 at 10:05 PM, stack <st...@duboce.net> wrote:
>>>>
>>>>
>>>>
>>>>> Then we should try and figure if there is an issue in the balancer, or
>>>>> maybe there is something missing if we are not doing a big upload in a
>>>>> manner that balances the upload across HDFS?
>>>>> St.Ack
>>>>>
>>>>> Slava Gorelik wrote:
>>>>>
>>>>>
>>>>>
>>>>>> Sure, i'll arrange logs tomorrow.About balancer, to wait when the
>>>>>> massive
>>>>>> work is finished is good in testing environment but in production it's
>>>>>> not
>>>>>> relevant :-)
>>>>>>
>>>>>> Best Regards.
>>>>>>
>>>>>> On Tue, Nov 4, 2008 at 9:48 PM, stack <st...@duboce.net> wrote:
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>> Slava Gorelik wrote:
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>> Hi.Regarding the failure of new block creation - i failed to run
>>>>>>>> hbase
>>>>>>>> till
>>>>>>>> i reformatted HDFS again.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>> I'd be interested in the logs.
>>>>>>>
>>>>>>> I just wandering if hadoop re balancing is necessary? Will it
>>>>>>> balance
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>> itself
>>>>>>>> ? As i understand hadoop balancer is moving data between data nodes,
>>>>>>>> but
>>>>>>>> in
>>>>>>>> my case this is during massive (8 clients just adding a records -
>>>>>>>> about
>>>>>>>> 400
>>>>>>>> requests for all region servers - 6). So, is it good idea to run
>>>>>>>> balancer during heavy load ?
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>> I don't have sufficient experience running the balancer. Perhaps
>>>>>>> wait
>>>>>>> till
>>>>>>> upload is done, then run it?
>>>>>>>
>>>>>>> St.Ack
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>
>>>
>>
>>
>
Re: Regionserver fails to serve region
Posted by Slava Gorelik <sl...@gmail.com>.
Thank You. Now it's clear.
On Tue, Nov 4, 2008 at 11:31 PM, stack <st...@duboce.net> wrote:
> Slava Gorelik wrote:
>
>> One more regarding the blockCache, how changes in store files (as i
>> understand those are MapFiles) are reflected on client side cache. If we
>> are
>> talking about more than one client that doing a changes ? If each client
>> has
>> different part of the MapFile ? or something else ?
>>
>>
>
> The block cache cache is over in the server. Its a cache for store files
> which never change once written. Did I say client-side cache? I should
> have been more clear. The client in this case is the regionserver itself.
> The cache is so the regionserver saves on its trips over the network
> visiting datanodes.
> St.Ack
>
>
>
> Best Regards.
>>
>> On Tue, Nov 4, 2008 at 11:10 PM, Slava Gorelik <slava.gorelik@gmail.com
>> >wrote:
>>
>>
>>
>>> I can try to reproduce it again, but before this i would like to send you
>>> a
>>> logs.
>>> Best Regards.
>>>
>>>
>>> On Tue, Nov 4, 2008 at 10:05 PM, stack <st...@duboce.net> wrote:
>>>
>>>
>>>
>>>> Then we should try and figure if there is an issue in the balancer, or
>>>> maybe there is something missing if we are not doing a big upload in a
>>>> manner that balances the upload across HDFS?
>>>> St.Ack
>>>>
>>>> Slava Gorelik wrote:
>>>>
>>>>
>>>>
>>>>> Sure, i'll arrange logs tomorrow.About balancer, to wait when the
>>>>> massive
>>>>> work is finished is good in testing environment but in production it's
>>>>> not
>>>>> relevant :-)
>>>>>
>>>>> Best Regards.
>>>>>
>>>>> On Tue, Nov 4, 2008 at 9:48 PM, stack <st...@duboce.net> wrote:
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>> Slava Gorelik wrote:
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>> Hi.Regarding the failure of new block creation - i failed to run
>>>>>>> hbase
>>>>>>> till
>>>>>>> i reformatted HDFS again.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>> I'd be interested in the logs.
>>>>>>
>>>>>> I just wandering if hadoop re balancing is necessary? Will it balance
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>> itself
>>>>>>> ? As i understand hadoop balancer is moving data between data nodes,
>>>>>>> but
>>>>>>> in
>>>>>>> my case this is during massive (8 clients just adding a records -
>>>>>>> about
>>>>>>> 400
>>>>>>> requests for all region servers - 6). So, is it good idea to run
>>>>>>> balancer during heavy load ?
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>> I don't have sufficient experience running the balancer. Perhaps wait
>>>>>> till
>>>>>> upload is done, then run it?
>>>>>>
>>>>>> St.Ack
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>
>>
>
>
Re: Regionserver fails to serve region
Posted by Slava Gorelik <sl...@gmail.com>.
Hi Michael.Did you have a chance to see the logs ?
Best Regards.
On Wed, Nov 5, 2008 at 1:24 PM, Slava Gorelik <sl...@gmail.com>wrote:
> Hi Michael.3'rd part (last) of the logs
>
>
> On Tue, Nov 4, 2008 at 11:31 PM, stack <st...@duboce.net> wrote:
>
>> Slava Gorelik wrote:
>>
>>> One more regarding the blockCache, how changes in store files (as i
>>> understand those are MapFiles) are reflected on client side cache. If we
>>> are
>>> talking about more than one client that doing a changes ? If each client
>>> has
>>> different part of the MapFile ? or something else ?
>>>
>>>
>>
>> The block cache cache is over in the server. Its a cache for store files
>> which never change once written. Did I say client-side cache? I should
>> have been more clear. The client in this case is the regionserver itself.
>> The cache is so the regionserver saves on its trips over the network
>> visiting datanodes.
>> St.Ack
>>
>>
>>
>> Best Regards.
>>>
>>> On Tue, Nov 4, 2008 at 11:10 PM, Slava Gorelik <slava.gorelik@gmail.com
>>> >wrote:
>>>
>>>
>>>
>>>> I can try to reproduce it again, but before this i would like to send
>>>> you a
>>>> logs.
>>>> Best Regards.
>>>>
>>>>
>>>> On Tue, Nov 4, 2008 at 10:05 PM, stack <st...@duboce.net> wrote:
>>>>
>>>>
>>>>
>>>>> Then we should try and figure if there is an issue in the balancer, or
>>>>> maybe there is something missing if we are not doing a big upload in a
>>>>> manner that balances the upload across HDFS?
>>>>> St.Ack
>>>>>
>>>>> Slava Gorelik wrote:
>>>>>
>>>>>
>>>>>
>>>>>> Sure, i'll arrange logs tomorrow.About balancer, to wait when the
>>>>>> massive
>>>>>> work is finished is good in testing environment but in production it's
>>>>>> not
>>>>>> relevant :-)
>>>>>>
>>>>>> Best Regards.
>>>>>>
>>>>>> On Tue, Nov 4, 2008 at 9:48 PM, stack <st...@duboce.net> wrote:
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>> Slava Gorelik wrote:
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>> Hi.Regarding the failure of new block creation - i failed to run
>>>>>>>> hbase
>>>>>>>> till
>>>>>>>> i reformatted HDFS again.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>> I'd be interested in the logs.
>>>>>>>
>>>>>>> I just wandering if hadoop re balancing is necessary? Will it
>>>>>>> balance
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>> itself
>>>>>>>> ? As i understand hadoop balancer is moving data between data nodes,
>>>>>>>> but
>>>>>>>> in
>>>>>>>> my case this is during massive (8 clients just adding a records -
>>>>>>>> about
>>>>>>>> 400
>>>>>>>> requests for all region servers - 6). So, is it good idea to run
>>>>>>>> balancer during heavy load ?
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>> I don't have sufficient experience running the balancer. Perhaps
>>>>>>> wait
>>>>>>> till
>>>>>>> upload is done, then run it?
>>>>>>>
>>>>>>> St.Ack
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>
>>>
>>
>>
>
Re: Regionserver fails to serve region
Posted by Slava Gorelik <sl...@gmail.com>.
Hi Michael.3'rd part (last) of the logs
On Tue, Nov 4, 2008 at 11:31 PM, stack <st...@duboce.net> wrote:
> Slava Gorelik wrote:
>
>> One more regarding the blockCache, how changes in store files (as i
>> understand those are MapFiles) are reflected on client side cache. If we
>> are
>> talking about more than one client that doing a changes ? If each client
>> has
>> different part of the MapFile ? or something else ?
>>
>>
>
> The block cache cache is over in the server. Its a cache for store files
> which never change once written. Did I say client-side cache? I should
> have been more clear. The client in this case is the regionserver itself.
> The cache is so the regionserver saves on its trips over the network
> visiting datanodes.
> St.Ack
>
>
>
> Best Regards.
>>
>> On Tue, Nov 4, 2008 at 11:10 PM, Slava Gorelik <slava.gorelik@gmail.com
>> >wrote:
>>
>>
>>
>>> I can try to reproduce it again, but before this i would like to send you
>>> a
>>> logs.
>>> Best Regards.
>>>
>>>
>>> On Tue, Nov 4, 2008 at 10:05 PM, stack <st...@duboce.net> wrote:
>>>
>>>
>>>
>>>> Then we should try and figure if there is an issue in the balancer, or
>>>> maybe there is something missing if we are not doing a big upload in a
>>>> manner that balances the upload across HDFS?
>>>> St.Ack
>>>>
>>>> Slava Gorelik wrote:
>>>>
>>>>
>>>>
>>>>> Sure, i'll arrange logs tomorrow.About balancer, to wait when the
>>>>> massive
>>>>> work is finished is good in testing environment but in production it's
>>>>> not
>>>>> relevant :-)
>>>>>
>>>>> Best Regards.
>>>>>
>>>>> On Tue, Nov 4, 2008 at 9:48 PM, stack <st...@duboce.net> wrote:
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>> Slava Gorelik wrote:
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>> Hi.Regarding the failure of new block creation - i failed to run
>>>>>>> hbase
>>>>>>> till
>>>>>>> i reformatted HDFS again.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>> I'd be interested in the logs.
>>>>>>
>>>>>> I just wandering if hadoop re balancing is necessary? Will it balance
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>> itself
>>>>>>> ? As i understand hadoop balancer is moving data between data nodes,
>>>>>>> but
>>>>>>> in
>>>>>>> my case this is during massive (8 clients just adding a records -
>>>>>>> about
>>>>>>> 400
>>>>>>> requests for all region servers - 6). So, is it good idea to run
>>>>>>> balancer during heavy load ?
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>> I don't have sufficient experience running the balancer. Perhaps wait
>>>>>> till
>>>>>> upload is done, then run it?
>>>>>>
>>>>>> St.Ack
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>
>>
>
>
Re: Regionserver fails to serve region
Posted by Slava Gorelik <sl...@gmail.com>.
Hi Michael.
Second part of the logs (will be 3 parts, due to mailing list limit).
On Tue, Nov 4, 2008 at 11:31 PM, stack <st...@duboce.net> wrote:
> Slava Gorelik wrote:
>
>> One more regarding the blockCache, how changes in store files (as i
>> understand those are MapFiles) are reflected on client side cache. If we
>> are
>> talking about more than one client that doing a changes ? If each client
>> has
>> different part of the MapFile ? or something else ?
>>
>>
>
> The block cache cache is over in the server. Its a cache for store files
> which never change once written. Did I say client-side cache? I should
> have been more clear. The client in this case is the regionserver itself.
> The cache is so the regionserver saves on its trips over the network
> visiting datanodes.
> St.Ack
>
>
>
> Best Regards.
>>
>> On Tue, Nov 4, 2008 at 11:10 PM, Slava Gorelik <slava.gorelik@gmail.com
>> >wrote:
>>
>>
>>
>>> I can try to reproduce it again, but before this i would like to send you
>>> a
>>> logs.
>>> Best Regards.
>>>
>>>
>>> On Tue, Nov 4, 2008 at 10:05 PM, stack <st...@duboce.net> wrote:
>>>
>>>
>>>
>>>> Then we should try and figure if there is an issue in the balancer, or
>>>> maybe there is something missing if we are not doing a big upload in a
>>>> manner that balances the upload across HDFS?
>>>> St.Ack
>>>>
>>>> Slava Gorelik wrote:
>>>>
>>>>
>>>>
>>>>> Sure, i'll arrange logs tomorrow.About balancer, to wait when the
>>>>> massive
>>>>> work is finished is good in testing environment but in production it's
>>>>> not
>>>>> relevant :-)
>>>>>
>>>>> Best Regards.
>>>>>
>>>>> On Tue, Nov 4, 2008 at 9:48 PM, stack <st...@duboce.net> wrote:
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>> Slava Gorelik wrote:
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>> Hi.Regarding the failure of new block creation - i failed to run
>>>>>>> hbase
>>>>>>> till
>>>>>>> i reformatted HDFS again.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>> I'd be interested in the logs.
>>>>>>
>>>>>> I just wandering if hadoop re balancing is necessary? Will it balance
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>> itself
>>>>>>> ? As i understand hadoop balancer is moving data between data nodes,
>>>>>>> but
>>>>>>> in
>>>>>>> my case this is during massive (8 clients just adding a records -
>>>>>>> about
>>>>>>> 400
>>>>>>> requests for all region servers - 6). So, is it good idea to run
>>>>>>> balancer during heavy load ?
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>> I don't have sufficient experience running the balancer. Perhaps wait
>>>>>> till
>>>>>> upload is done, then run it?
>>>>>>
>>>>>> St.Ack
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>
>>
>
>
Re: Regionserver fails to serve region
Posted by stack <st...@duboce.net>.
Slava Gorelik wrote:
> One more regarding the blockCache, how changes in store files (as i
> understand those are MapFiles) are reflected on client side cache. If we are
> talking about more than one client that doing a changes ? If each client has
> different part of the MapFile ? or something else ?
>
The block cache cache is over in the server. Its a cache for store files
which never change once written. Did I say client-side cache? I should
have been more clear. The client in this case is the regionserver
itself. The cache is so the regionserver saves on its trips over the
network visiting datanodes.
St.Ack
> Best Regards.
>
> On Tue, Nov 4, 2008 at 11:10 PM, Slava Gorelik <sl...@gmail.com>wrote:
>
>
>> I can try to reproduce it again, but before this i would like to send you a
>> logs.
>> Best Regards.
>>
>>
>> On Tue, Nov 4, 2008 at 10:05 PM, stack <st...@duboce.net> wrote:
>>
>>
>>> Then we should try and figure if there is an issue in the balancer, or
>>> maybe there is something missing if we are not doing a big upload in a
>>> manner that balances the upload across HDFS?
>>> St.Ack
>>>
>>> Slava Gorelik wrote:
>>>
>>>
>>>> Sure, i'll arrange logs tomorrow.About balancer, to wait when the massive
>>>> work is finished is good in testing environment but in production it's
>>>> not
>>>> relevant :-)
>>>>
>>>> Best Regards.
>>>>
>>>> On Tue, Nov 4, 2008 at 9:48 PM, stack <st...@duboce.net> wrote:
>>>>
>>>>
>>>>
>>>>
>>>>> Slava Gorelik wrote:
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>> Hi.Regarding the failure of new block creation - i failed to run hbase
>>>>>> till
>>>>>> i reformatted HDFS again.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>> I'd be interested in the logs.
>>>>>
>>>>> I just wandering if hadoop re balancing is necessary? Will it balance
>>>>>
>>>>>
>>>>>
>>>>>> itself
>>>>>> ? As i understand hadoop balancer is moving data between data nodes,
>>>>>> but
>>>>>> in
>>>>>> my case this is during massive (8 clients just adding a records - about
>>>>>> 400
>>>>>> requests for all region servers - 6). So, is it good idea to run
>>>>>> balancer during heavy load ?
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>> I don't have sufficient experience running the balancer. Perhaps wait
>>>>> till
>>>>> upload is done, then run it?
>>>>>
>>>>> St.Ack
>>>>>
>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>
>
Re: Regionserver fails to serve region
Posted by Slava Gorelik <sl...@gmail.com>.
One more regarding the blockCache, how changes in store files (as i
understand those are MapFiles) are reflected on client side cache. If we are
talking about more than one client that doing a changes ? If each client has
different part of the MapFile ? or something else ?
Best Regards.
On Tue, Nov 4, 2008 at 11:10 PM, Slava Gorelik <sl...@gmail.com>wrote:
> I can try to reproduce it again, but before this i would like to send you a
> logs.
> Best Regards.
>
>
> On Tue, Nov 4, 2008 at 10:05 PM, stack <st...@duboce.net> wrote:
>
>> Then we should try and figure if there is an issue in the balancer, or
>> maybe there is something missing if we are not doing a big upload in a
>> manner that balances the upload across HDFS?
>> St.Ack
>>
>> Slava Gorelik wrote:
>>
>>> Sure, i'll arrange logs tomorrow.About balancer, to wait when the massive
>>> work is finished is good in testing environment but in production it's
>>> not
>>> relevant :-)
>>>
>>> Best Regards.
>>>
>>> On Tue, Nov 4, 2008 at 9:48 PM, stack <st...@duboce.net> wrote:
>>>
>>>
>>>
>>>> Slava Gorelik wrote:
>>>>
>>>>
>>>>
>>>>> Hi.Regarding the failure of new block creation - i failed to run hbase
>>>>> till
>>>>> i reformatted HDFS again.
>>>>>
>>>>>
>>>>>
>>>>>
>>>> I'd be interested in the logs.
>>>>
>>>> I just wandering if hadoop re balancing is necessary? Will it balance
>>>>
>>>>
>>>>> itself
>>>>> ? As i understand hadoop balancer is moving data between data nodes,
>>>>> but
>>>>> in
>>>>> my case this is during massive (8 clients just adding a records - about
>>>>> 400
>>>>> requests for all region servers - 6). So, is it good idea to run
>>>>> balancer during heavy load ?
>>>>>
>>>>>
>>>>>
>>>>>
>>>> I don't have sufficient experience running the balancer. Perhaps wait
>>>> till
>>>> upload is done, then run it?
>>>>
>>>> St.Ack
>>>>
>>>>
>>>>
>>>
>>>
>>>
>>
>>
>
Re: Regionserver fails to serve region
Posted by Slava Gorelik <sl...@gmail.com>.
I can try to reproduce it again, but before this i would like to send you a
logs.
Best Regards.
On Tue, Nov 4, 2008 at 10:05 PM, stack <st...@duboce.net> wrote:
> Then we should try and figure if there is an issue in the balancer, or
> maybe there is something missing if we are not doing a big upload in a
> manner that balances the upload across HDFS?
> St.Ack
>
> Slava Gorelik wrote:
>
>> Sure, i'll arrange logs tomorrow.About balancer, to wait when the massive
>> work is finished is good in testing environment but in production it's not
>> relevant :-)
>>
>> Best Regards.
>>
>> On Tue, Nov 4, 2008 at 9:48 PM, stack <st...@duboce.net> wrote:
>>
>>
>>
>>> Slava Gorelik wrote:
>>>
>>>
>>>
>>>> Hi.Regarding the failure of new block creation - i failed to run hbase
>>>> till
>>>> i reformatted HDFS again.
>>>>
>>>>
>>>>
>>>>
>>> I'd be interested in the logs.
>>>
>>> I just wandering if hadoop re balancing is necessary? Will it balance
>>>
>>>
>>>> itself
>>>> ? As i understand hadoop balancer is moving data between data nodes, but
>>>> in
>>>> my case this is during massive (8 clients just adding a records - about
>>>> 400
>>>> requests for all region servers - 6). So, is it good idea to run
>>>> balancer during heavy load ?
>>>>
>>>>
>>>>
>>>>
>>> I don't have sufficient experience running the balancer. Perhaps wait
>>> till
>>> upload is done, then run it?
>>>
>>> St.Ack
>>>
>>>
>>>
>>
>>
>>
>
>
Re: Regionserver fails to serve region
Posted by stack <st...@duboce.net>.
Then we should try and figure if there is an issue in the balancer, or
maybe there is something missing if we are not doing a big upload in a
manner that balances the upload across HDFS?
St.Ack
Slava Gorelik wrote:
> Sure, i'll arrange logs tomorrow.About balancer, to wait when the massive
> work is finished is good in testing environment but in production it's not
> relevant :-)
>
> Best Regards.
>
> On Tue, Nov 4, 2008 at 9:48 PM, stack <st...@duboce.net> wrote:
>
>
>> Slava Gorelik wrote:
>>
>>
>>> Hi.Regarding the failure of new block creation - i failed to run hbase
>>> till
>>> i reformatted HDFS again.
>>>
>>>
>>>
>> I'd be interested in the logs.
>>
>> I just wandering if hadoop re balancing is necessary? Will it balance
>>
>>> itself
>>> ? As i understand hadoop balancer is moving data between data nodes, but
>>> in
>>> my case this is during massive (8 clients just adding a records - about
>>> 400
>>> requests for all region servers - 6). So, is it good idea to run
>>> balancer during heavy load ?
>>>
>>>
>>>
>> I don't have sufficient experience running the balancer. Perhaps wait till
>> upload is done, then run it?
>>
>> St.Ack
>>
>>
>
>
Re: Regionserver fails to serve region
Posted by Slava Gorelik <sl...@gmail.com>.
Sure, i'll arrange logs tomorrow.About balancer, to wait when the massive
work is finished is good in testing environment but in production it's not
relevant :-)
Best Regards.
On Tue, Nov 4, 2008 at 9:48 PM, stack <st...@duboce.net> wrote:
> Slava Gorelik wrote:
>
>> Hi.Regarding the failure of new block creation - i failed to run hbase
>> till
>> i reformatted HDFS again.
>>
>>
> I'd be interested in the logs.
>
> I just wandering if hadoop re balancing is necessary? Will it balance
>> itself
>> ? As i understand hadoop balancer is moving data between data nodes, but
>> in
>> my case this is during massive (8 clients just adding a records - about
>> 400
>> requests for all region servers - 6). So, is it good idea to run
>> balancer during heavy load ?
>>
>>
> I don't have sufficient experience running the balancer. Perhaps wait till
> upload is done, then run it?
>
> St.Ack
>
Re: Regionserver fails to serve region
Posted by stack <st...@duboce.net>.
Slava Gorelik wrote:
> Hi.Regarding the failure of new block creation - i failed to run hbase till
> i reformatted HDFS again.
>
I'd be interested in the logs.
> I just wandering if hadoop re balancing is necessary? Will it balance itself
> ? As i understand hadoop balancer is moving data between data nodes, but in
> my case this is during massive (8 clients just adding a records - about 400
> requests for all region servers - 6). So, is it good idea to run
> balancer during heavy load ?
>
I don't have sufficient experience running the balancer. Perhaps wait
till upload is done, then run it?
St.Ack
Re: Regionserver fails to serve region
Posted by Slava Gorelik <sl...@gmail.com>.
Hi.Regarding the failure of new block creation - i failed to run hbase till
i reformatted HDFS again.
I just wandering if hadoop re balancing is necessary? Will it balance itself
? As i understand hadoop balancer is moving data between data nodes, but in
my case this is during massive (8 clients just adding a records - about 400
requests for all region servers - 6). So, is it good idea to run
balancer during heavy load ?
Best Regards.
On Tue, Nov 4, 2008 at 9:33 PM, stack <st...@duboce.net> wrote:
> Slava Gorelik wrote:
>
>> Hi.I happened yesterday after 28 hours of running i started the hadoop
>> balancer after 2 hours working (with some exception- ca't move block) my
>> hbase started to throws exception that can't create new block. It happened
>> on on couple of region servers and then master is failed to connect to
>> then
>> and eventually it crashed. Tomorrow I can cut last 2-3 hours from logs
>> (they
>> are huge) and send you.
>>
>>
>
> Sorry about that. I presumed it safe given that balancer has been around a
> few releases and we're running it here continuously w/o issue. Did estart
> fix things or are there now missing blocks?
>
>> BTW, some i sent email to list couple of days ago about blockCache
>> parameter
>> on column family descriptor, what is it and how it affect on performance ?
>>
>>
> Sorry, missed it.
>
> The blockcache is client-side caching of pieces of store files. You can
> set the size of the blocks to cache client-side. It uses java Soft
> References. Blocks are evicted on roughly an LRU basis when memory is low.
> It was added a good while ago by Tom White.
>
> By default it has been off but as of HBASE-953 commit of about a week or so
> ago, after some playing and tuning, the default has been flipped and now
> blockcache is on by default. Block caching along with other performance
> improvements including rpc fixes and J-D's scanner pre-fetching and batch
> writing, will make the 0.19.0 release run faster than its predecessors in
> many regards.
>
> Some rough benchmarking running our performance test -- keep in mind, this
> is not-very-real-world just a single client going against a single
> regionserver (see wiki for more) -- shows writes running at ~3X speed, scans
> at ~7X, sequential reads at ~2X and random reads anywhere from slower to 2
> to 3 times faster dependent on how well the block cache is helping (or
> hindering). If the regionserver has more memory, random reads run faster.
> If not enough, regionserver is just spinning filling cache and random read
> times plummet. I'll put up some numbers when we come closer to the 0.19.0
> release.
>
> If you do enable block cache, be sure to update your hbase-default.xml.
> The old block size tends to provoke OOMEs.
> St.Ack
>
Re: Regionserver fails to serve region
Posted by stack <st...@duboce.net>.
Slava Gorelik wrote:
> Hi.I happened yesterday after 28 hours of running i started the hadoop
> balancer after 2 hours working (with some exception- ca't move block) my
> hbase started to throws exception that can't create new block. It happened
> on on couple of region servers and then master is failed to connect to then
> and eventually it crashed. Tomorrow I can cut last 2-3 hours from logs (they
> are huge) and send you.
>
Sorry about that. I presumed it safe given that balancer has been
around a few releases and we're running it here continuously w/o issue.
Did estart fix things or are there now missing blocks?
> BTW, some i sent email to list couple of days ago about blockCache parameter
> on column family descriptor, what is it and how it affect on performance ?
>
Sorry, missed it.
The blockcache is client-side caching of pieces of store files. You can
set the size of the blocks to cache client-side. It uses java Soft
References. Blocks are evicted on roughly an LRU basis when memory is
low. It was added a good while ago by Tom White.
By default it has been off but as of HBASE-953 commit of about a week or
so ago, after some playing and tuning, the default has been flipped and
now blockcache is on by default. Block caching along with other
performance improvements including rpc fixes and J-D's scanner
pre-fetching and batch writing, will make the 0.19.0 release run faster
than its predecessors in many regards.
Some rough benchmarking running our performance test -- keep in mind,
this is not-very-real-world just a single client going against a single
regionserver (see wiki for more) -- shows writes running at ~3X speed,
scans at ~7X, sequential reads at ~2X and random reads anywhere from
slower to 2 to 3 times faster dependent on how well the block cache is
helping (or hindering). If the regionserver has more memory, random
reads run faster. If not enough, regionserver is just spinning filling
cache and random read times plummet. I'll put up some numbers when we
come closer to the 0.19.0 release.
If you do enable block cache, be sure to update your hbase-default.xml.
The old block size tends to provoke OOMEs.
St.Ack
Re: Regionserver fails to serve region
Posted by Slava Gorelik <sl...@gmail.com>.
Hi.I happened yesterday after 28 hours of running i started the hadoop
balancer after 2 hours working (with some exception- ca't move block) my
hbase started to throws exception that can't create new block. It happened
on on couple of region servers and then master is failed to connect to then
and eventually it crashed. Tomorrow I can cut last 2-3 hours from logs (they
are huge) and send you.
Thank You for Your assistance.
BTW, some i sent email to list couple of days ago about blockCache parameter
on column family descriptor, what is it and how it affect on performance ?
Best Regards.
On Tue, Nov 4, 2008 at 7:27 PM, stack <st...@duboce.net> wrote:
> Did it? It shouldn't. If it does, then its a bug we need to figure.
> Whats its log say? Otherwise you could do the rebalance manually but your
> results will be spotter (Up replication, let it run a while, then
> selectively shutdown nodes checking filesystem as you go to make sure at
> least one replica is still being served and then restore original
> replication level, etc.).
> St.Ack
>
> Slava Gorelik wrote:
>
>> Can i run balancer on hadoop during massive load on Hbase ?Last time i did
>> it i killed my data :-)
>>
>> On Tue, Nov 4, 2008 at 10:45 AM, Michael Stack <st...@duboce.net> wrote:
>>
>>
>>
>>> Slava Gorelik wrote:
>>>
>>>
>>>
>>>> Hi Michael.After reformatting HDFS, Hbase started to work as a Swiss
>>>> Clock.
>>>> Worked with 8 clients about 30 hours intensive load.
>>>>
>>>>
>>>>
>>>>
>>> Thanks for reporting back to the list.
>>>
>>>
>>>
>>>> Just small question, after about 28 hours (when i came back to work) i
>>>> found
>>>> that one of 7 datanodes in Hadoop is about 98% usage and all other about
>>>> 30%, is it normal ?
>>>>
>>>>
>>>>
>>>>
>>> I haven't kept a close eye on HDFS usage during hbase upload. I do know
>>> that out-of-balance would seem to be a common condition and that its been
>>> reported on the list that hbase runs faster on a balanced HDFS. You
>>> might
>>> try out the hadoop balancer. There's a little note on it here in the
>>> hbase
>>> FAQ: http://wiki.apache.org/hadoop/Hbase/FAQ#7.
>>>
>>> St.Ack
>>>
>>>
>>>
>>
>>
>>
>
>
Re: Regionserver fails to serve region
Posted by stack <st...@duboce.net>.
Did it? It shouldn't. If it does, then its a bug we need to figure.
Whats its log say? Otherwise you could do the rebalance manually but
your results will be spotter (Up replication, let it run a while, then
selectively shutdown nodes checking filesystem as you go to make sure at
least one replica is still being served and then restore original
replication level, etc.).
St.Ack
Slava Gorelik wrote:
> Can i run balancer on hadoop during massive load on Hbase ?Last time i did
> it i killed my data :-)
>
> On Tue, Nov 4, 2008 at 10:45 AM, Michael Stack <st...@duboce.net> wrote:
>
>
>> Slava Gorelik wrote:
>>
>>
>>> Hi Michael.After reformatting HDFS, Hbase started to work as a Swiss
>>> Clock.
>>> Worked with 8 clients about 30 hours intensive load.
>>>
>>>
>>>
>> Thanks for reporting back to the list.
>>
>>
>>> Just small question, after about 28 hours (when i came back to work) i
>>> found
>>> that one of 7 datanodes in Hadoop is about 98% usage and all other about
>>> 30%, is it normal ?
>>>
>>>
>>>
>> I haven't kept a close eye on HDFS usage during hbase upload. I do know
>> that out-of-balance would seem to be a common condition and that its been
>> reported on the list that hbase runs faster on a balanced HDFS. You might
>> try out the hadoop balancer. There's a little note on it here in the hbase
>> FAQ: http://wiki.apache.org/hadoop/Hbase/FAQ#7.
>>
>> St.Ack
>>
>>
>
>
Re: Regionserver fails to serve region
Posted by Slava Gorelik <sl...@gmail.com>.
Can i run balancer on hadoop during massive load on Hbase ?Last time i did
it i killed my data :-)
On Tue, Nov 4, 2008 at 10:45 AM, Michael Stack <st...@duboce.net> wrote:
> Slava Gorelik wrote:
>
>> Hi Michael.After reformatting HDFS, Hbase started to work as a Swiss
>> Clock.
>> Worked with 8 clients about 30 hours intensive load.
>>
>>
> Thanks for reporting back to the list.
>
>> Just small question, after about 28 hours (when i came back to work) i
>> found
>> that one of 7 datanodes in Hadoop is about 98% usage and all other about
>> 30%, is it normal ?
>>
>>
> I haven't kept a close eye on HDFS usage during hbase upload. I do know
> that out-of-balance would seem to be a common condition and that its been
> reported on the list that hbase runs faster on a balanced HDFS. You might
> try out the hadoop balancer. There's a little note on it here in the hbase
> FAQ: http://wiki.apache.org/hadoop/Hbase/FAQ#7.
>
> St.Ack
>
Re: Regionserver fails to serve region
Posted by Michael Stack <st...@duboce.net>.
Slava Gorelik wrote:
> Hi Michael.After reformatting HDFS, Hbase started to work as a Swiss Clock.
> Worked with 8 clients about 30 hours intensive load.
>
Thanks for reporting back to the list.
> Just small question, after about 28 hours (when i came back to work) i found
> that one of 7 datanodes in Hadoop is about 98% usage and all other about
> 30%, is it normal ?
>
I haven't kept a close eye on HDFS usage during hbase upload. I do know
that out-of-balance would seem to be a common condition and that its
been reported on the list that hbase runs faster on a balanced HDFS.
You might try out the hadoop balancer. There's a little note on it here
in the hbase FAQ: http://wiki.apache.org/hadoop/Hbase/FAQ#7.
St.Ack